Discussion Summary: Is the European Union's AI Policy Steering Correctly?
The European Commission has published more details about its public consultation on the white paper on Artificial Intelligence (AI), revealing insights into the EU's approach to AI development and regulation.
A significant finding from the consultation is the attraction of the U.S. market for EU AI start-ups and scale-ups, with the U.S. market being perceived as more appealing than the fragmented EU ecosystem. However, the GDPR's obstacles in access to personal data for AI development remain a critical point of discussion.
The EU's vision for AI, based on fundamental rights, contrasts with China's approach. The EU can offer an alternative, including through its industrial strategy, a human-centric approach, and global collaboration. The EU's privacy law, the GDPR, was written before the AI wave, and a legal framework that lacks pragmatism and agility may prevent EU start-ups from growing and competing internationally.
To address these concerns, 85 percent of respondents favor ex-ante assessment measures regarding AI, and a gradual approach to future AI legislation is advocated, with alignment with the GDPR as the law covers a few aspects of automated decision making. The EU and member states should consider relaxing some of the GDPR's binding rules that restrict the collection and use of personal data for AI systems to foster innovation.
The consultation's responses recommend that the EU closely cooperate with member states in implementing its AI approach. Most respondents believe that regulations should cover biometric tools regarding AI, and 43 percent of respondents say that compulsory requirements should apply only to high-risk AI.
Data protection is key to AI development, but stringent regulation may limit AI and machine learning applications in the EU. The Commission is planning a virtual event for the European AI Alliance at the beginning of October to further discuss these issues.
China's ambitious AI strategy and increasing involvement in standard-setting bodies is a concern for the EU. The EU single digital market is not fully integrated yet, and this may be a challenge for the EU in leading in AI.
Germany and Austria have supported a learning-oriented and pragmatic strategy for AI development, emphasizing clear legal frameworks that enable innovation and practical testing through regulatory sandboxes. They plan to mutually support the implementation of EU AI approaches through close European coordination, national market surveillance authorities, and collaborative platforms like Gaia-X to foster interoperable federated data spaces and data sovereignty.
After analyzing the public feedback, the Commission will release a legislative proposal on AI. Of the 1,215 responses to the consultation, only 3 percent say that current EU legislation sufficiently addresses concerns regarding AI. 33 percent of respondents favor changing existing legislation regarding AI.
Regulation should support, rather than obstruct, the benefits AI can deliver for innovation, society, and economic growth. The EU should create environments allowing innovators to make mistakes and experiment with AI applications, such as through digital innovation hubs and fields labs.
In conclusion, the EU's approach to AI development and regulation is undergoing significant changes, with a focus on fostering innovation, cooperation, and pragmatism. The Commission's forthcoming legislative proposal on AI will be a crucial step in shaping the EU's future in AI.
